gene repair of an usher syndrome causing mutation by zinc-finger nuclease … · 2019-01-10 ·...

Genetics

Gene Repair of an Usher Syndrome CausingMutation by Zinc-Finger Nuclease MediatedHomologous Recombination

Nora Overlack, Tobias Goldmann,* Uwe Wolfrum,1 and Kerstin Nagel-Wolfrum1

PURPOSE. Human Usher syndrome (USH) is the most frequentcause of inherited deaf-blindness. It is clinically and geneticallyheterogeneous, assigned to three clinical types of which themost severe type is USH1. No effective treatment for theophthalmic component of USH exists. Gene augmentation isan attractive strategy for hereditary retinal diseases. However,several USH genes, like USH1C, are expressed in variousisoforms, hampering gene augmentation. As an alternativetreatment strategy, we applied the zinc-finger nuclease (ZFN)technology for targeted gene repair of an USH1C, causingmutation by homologous recombination.

METHODS. We designed ZFNs customized for the p.R31Xnonsense mutation in Ush1c. We evaluated ZFNs for DNAcleavage capability and analyzed ZFNs biocompatibilities byXTT assays. We demonstrated ZFNs mediated gene repair ongenomic level by digestion assays and DNA sequencing, and onprotein level by indirect immunofluorescence and Westernblot analyses.

RESULTS. The specifically designed ZFNs did not show cytotoxiceffects in a p.R31X cell line. We demonstrated that ZFNinduced cleavage of their target sequence. We showed thatsimultaneous application of ZFN and rescue DNA induced generepair of the disease-causing mutation on the genomic level,resulting in recovery of protein expression.

CONCLUSIONS. In our present study, we analyzed for the first timeZFN-activated gene repair of an USH gene. The data highlightthe ability of ZFNs to induce targeted homologous recombina-tion and mediate gene repair in USH. We provide furtherevidence that the ZFN technology holds great potential torecover disease-causing mutations in inherited retinal disor-ders. (Invest Ophthalmol Vis Sci. 2012;53:4140–4146) DOI:10.1167/iovs.12-9812

The human Usher syndrome (USH) is the most frequentcause of inherited combined deaf-blindness. USH is

clinically and genetically heterogeneous and is assigned tothree clinical USH types.1 The most severe form of the diseaseis USH1, characterized by profound prelingual hearing loss,vestibular areflexia, and prepubertal onset of retinitis pigmen-

tosa (RP).1,2 Five USH1 genes (USH1B–G) and their geneproducts have been identified.1,2 In the present study, wefocused on the USH1C gene coding for the scaffold proteinharmonin.3 Harmonin was previously identified as one of thekey organizers in the USH protein interactome.1,2 So far noeffective clinical treatment for the ophthalmic component ofUSH exists.

Gene-based strategies became attractive for the treatment ofhereditary retinal degenerations (e.g., gene augmentation byadeno-associated viruses for the treatment of Leber CongenitalAmaurosis [LCA]).4 However, USH is a difficult target for viralgene augmentation approaches; several USH genes exceed thepackaging size of therapeutic viruses. In addition, thetranscripts of most USH genes are alternatively spliced, andseveral splice variant isoforms have to be added for anaugmentative treatment.5 In particular, recent expressionanalysis of USH1C in human retinas demonstrated thepresence of several harmonin transcripts.6 The ambiguityconcerning the activities of different isoforms hampers viralgene augmentation approaches for USH1C patients.

The most elegant way to treat a genetic disorder is torepair the disease-causing mutation in its endogenous locus.7

In contrast to other gene-based therapy approaches, in thismethod, the expression and splicing of the corrected generemains under endogenous control, the size of the gene is noissue, and the type of the disease-causing mutation does nothave to be taken into account. Gene repair, also calledgenome editing or gene correction, can be achieved byhomologous recombination (HR) within the cell. Unfortu-nately, HR occurs under normal conditions at a very lowfrequency of 10�6 and therefore cannot be used fortherapeutic means.8,9 However, the frequency of HR can beincreased several thousand-fold by the introduction of double-strand breaks (DSBs) in the DNA.10 Zinc-finger nucleases(ZFNs) are engineered genetic tools designed to specificallycreate DSBs at a preselected genomic target sequence andthereby activate HR.11,12 ZFN modules are chimeric proteinscomposed of a DNA-binding domain (ZF domain) fused to thenonspecific nuclease domain of FokI (Fig. 1).7 The ZFdomains can be specifically designed to bind to almost anypreselected DNA sequence in the genome.12 To introduce aDSB, two ZFN modules have to form a dimer in order toactivate the FokI nuclease activity (Fig. 1).7 This requires twoZFN modules, binding opposing targets across a spacer wherethe FokI domains come together to create the break (Fig.1).7,9 In a ZFN module, each zinc-finger subunit (Fig. 1, grey

From the Institute of Zoology, Department of Cell and MatrixBiology, Johannes Gutenberg University of Mainz, Mainz, Germany.

1These authors contributed equally to this work and shouldtherefore be regarded as equivalent authors.

Supported by grants from the Deutsche Forschungsgemein-schaft (GRK 1044), FAUN-Stiftung, Nuremberg, and by the Founda-tion Fighting Blindness (FFB).

Submitted for publication March 7, 2012; revised April 30,2012; accepted May 22, 2012.

Disclosure: N. Overlack, None; T. Goldmann, None; U.Wolfrum, None; K. Nagel-Wolfrum, None

Current affiliation: *Medical School Freiburg, Institute of Pathol-ogy, Department for Neuropathology, Freiburg, Germany.

Corresponding author: Kerstin Nagel-Wolfrum, Johannes Guten-berg University, Institute of Zoology, Department of Cell and MatrixBiology, D-55099 Mainz, Germany; [email protected].

Investigative Ophthalmology & Visual Science, June 2012, Vol. 53, No. 7

4140 Copyright 2012 The Association for Research in Vision and Ophthalmology, Inc.

ovals; ZF) recognizes and binds to 3 base pairs (bp). Cloningof three ZFs in line results in binding to nine consecutive basepairs. Dimerization of two ZFN modules therefore recognizes18 bp, making binding of the ZFN to DNA strands a likelyunique event.7,11 To repair the DSB during HR, an exoge-nously introduced DNA rescue plasmid, containing thedesired sequence, can be used as a template.13–15

Here we designed customized ZFNs, specific for the p.R31Xmutation of murine Ush1c. We show the activity of ZFN bycleavage of the DNA target sequence. In addition, wedemonstrate gene correction of the p.R31X mutation on thegenomic level and expression of the repaired gene product,harmonin. These results highlight the capacity of ZFNs to treatindividuals with mutations in USH1C and other inheritedocular and nonocular diseases as well.

MATERIALS AND METHODS

Design and Generation of ZFN for Ush1C-P.r31XMutation

The mutated Ush1c sequence was screened for putative ZF binding

sites with the Zinc-Finger consortium–based software ZiFiT (in the

public domain; http://bindr.gdcb.iastate.edu/ZiFiT).16 Cloning of ZFN

by modular assembly was performed as previously described.17 ZFN

consisted of three individual ZF-DNA binding sites fused to the FokI

domain (Addgene plasmid 13,426). Resulting ZFN modules were each

cloned into an eukaryotic expression vector (Addgene plasmid

pST1374) with a nuclear localization signal (NLS) and a Flag-tag to

allow detection of the ZFN module.

P.r31X-Ush1c Stable Cell Line and Rescue Plasmid

The p.R31X mutation was introduced into mouse Ush1c cDNA by PCR

site-directed mutagenesis applying the QuickChange Lightning Site-

Directed Mutagenesis Kit (Stratgene, La Jolla, CA). The Ush1c and

mutated cDNA (bp 1-999) was subcloned in a pDEST expression vector

with a C-terminal Myc-tag. Stable p.R31X-Ush1c or Ush1c cell lines

were generated by transfecting plasmids containing either p.R31X-

Ush1c or Ush1c cDNA regulated by a cytomegalovirus (CMV) promoter

into Flp in-HEK293 cells and selection by hygromycin B according to

manufacturer’s protocol (Invitrogen, Karlsruhe, Germany). The rescue

plasmid contained the wild type Ush1c cDNA (bp 1-999). The

sequence was cloned into pGL3-Basic vector (Promega, Fitchburg,

WI), which has no promoter.

Cell Culture

Multi-clone p.R31X-Ush1c and Ush1c cell lines were grown in

Dulbecco’s modified Eagle’s medium (DMEM) with GlutaMax

supplemented with 10% fetal calf serum (FCS; Invitrogen), 1%

penicillin/streptomycin (Invitrogen) at 378C, and 5% CO2. Transfec-

tions were performed with Lipofectamine LTX and PLUS reagent

(Invitrogen) according to manufacturer’s protocol. After 6 hours,

medium was changed to fresh medium, and cells were grown for 48

to 72 hours.

Assessment of ZFN-Related Cytotoxicity

Stable p.R31X cells were transfected with ZFNs. Forty-eight hours post

transfection, ZFN toxicity was assessed by XTT cell viability assay

(Roche, Duren, Germany) according to manufacturer’s instructions.

Assays were performed as triplets. Presented data are an average of five

to nine independent experiments. G418 (1 mg/mL) was used as

positive control for the assay.

Semiquantitative PCR

Forty-eight-hours post transfection of stable p.R31X cells with ZFNs,

genomic DNA was isolated by the High Pure PCR Template Preparation

Kit (Roche). As positive control for the cleavage, genomic DNA of

stable p.R31X cells was digested with EcoRI. Duplex PCRs were

performed in a volume of 50 lL using 12 ng genomic DNA and 10 nmol

of each primer/reaction. Cycling conditions were 25 cycles at 948C for

40 seconds, 608C for 40 seconds, and 728C for 1 minute followed by a

7-minute 728C extension. Harmonin-specific primers flanking the

potential cleavage site (forward 50-ATGGACCGGAAGGTGGCCCGA-30

and reverse primer 50 CCTTCTGGGTGCAGACGGTCCAAG-30) were

used. Primer set for GAPDH (forward 50-TGCACCACCAACTGCTTAGC-

30 and reverse 50-GGCATGGACTGTGGTCATGAG-3 0) was used as

internal control for the PCR reaction. PCR products were analyzed

on ethidium bromide–stained 1% agarose gels.

FIGURE 1. Zinc-finger nucleases for the Ush1c p.R31X mutation. (A)Zinc-finger nucleases are chimeric proteins consisting of DNA bindingzinc-fingers (ZF) fused to the nonspecific cleaving endonuclease FokI(ZFN module). We designed ZFN modules with three DNA bindingZFs (1–3 grey ovals, ZF domain). The FokI is only functional whendimerized, making a pair of ZFN modules necessary to form an activeZFN heterodimer and introduce a DNA double-strand break. Segmentsfrom bp 74 to 96 and bp 87 to 109 were chosen because one ZF isbinding directly to a mutated base pair (asterisks). (B) In silicoscreens resulted in seven different potential ZFN modules for p.R31X(A–G).

IOVS, June 2012, Vol. 53, No. 7 USH1C Gene Repair 4141

TseI Digest

Seventy-two hours post transfection, genomic DNA was isolated from

cells transfected with ZFNs and rescue plasmid by the High Pure PCR

Template Preparation Kit (Roche). PCRs were performed in a volume

of 50 lL, using 300 ng prepared genomic DNA and 10 nmol of each

primer per reaction. Cycling conditions were 35 cycles at 948C for 40

seconds, 608C for 40 seconds, and 728C for 1 minute followed by a 7-

minute 728C extension. Forward primer binding in the CMV-promoter

of stable transfected p.R31X construct (50-GTACGGTGGGAGGTCTA-

TAT-30) and a harmonin-specific reverse primer (50 CCTTCTGGGTGCA-

GACGGTCCAAG-30) were used. Ten microliters PCR product were

digested with 0.5 lL TseI (New England Biolabs, Ipswich, MA) for 2

hours at 658C and digestion analyzed on ethidium bromide–stained 2%

agarose gels.

DNA Sequencing

Sequencing of 500 bp PCR products after gene repair was performed

by StarSeq (Mainz, Germany). Sequencing results were analyzed, and

DNA alignment performed with BioEdit Version 7 (Carlsbad, CA).

Antibodies and Dyes

Monoclonal mouse antibodies to a-tubulin (clone DM1A), Flag (M2)

were obtained from Sigma-Aldrich (St. Louis, MO) and Myc (clone

9B11) from Cell Signaling (Danvers, MA). Polyclonal rabbit antibodies

against harmonin (H3) were used as previously described.18 Secondary

antibodies conjugated to Alexa 488 or 568 were purchased from

Molecular Probes (Leiden, The Netherlands) and DAPI (40,6-Diamidino-

2-phenylindol) from Sigma-Aldrich. For Western blot, secondary

antibodies were purchased from Invitrogen or Rockland (Gilbertsville,

PA).

Immunofluorescence Microscopy and WesternBlot Analysis

Immunofluorescence analyses were carried out on stable transfected

p.R31X cells seeded on coverslips, followed by fixation with methanol,

and proceeded as previously described.19 Specimens were analyzed in

a Leica DM 6000 B microscope (Leica Microsystems, Bensheim,

Germany) and processed with Adobe Photoshop CS (Adobe Systems,

San Jose, CA). Harvesting of cells and Western blot analyses were

performed as previously described.20,21

RESULTS

Design of Specific ZFNs for a Nonsense Mutation inUsh1C

In order to design specific ZFNs for target sites in the relevantmutation in USH1C, we used the ZiFiT software.16 Wescreened the sequence coding for the murine Ush1c ortho-logue, which contains an USH1 disease-causing p.R31Xnonsense mutation. In this mutation, the wild type CGAarginine codon (R) at position 91 is altered to a premature stopcodon TGA (X). For the ZFN design, using ZiFiT, we appliedsoftware settings for ZF modules, which recognize a nine-bpsequence at the putative cleavage site separated by a spacer offive bp (Fig. 1A). Based on these prerequisites, we identified bp74 to 96 and bp 87 to 109 as targets in the mutated Ush1c

coding sequence for the design of promising specific ZFNs(Figs. 1A, B). We designed the ZF domains in order that one ZFdirectly binds to mutated base pairs. We generated a total ofseven predicted ZF domains by modular assembly, which werefused with the FokI nuclease domain to form customized ZFNmodules.16,17 This strategy resulted in six matching ZFNs (A/C,A/D, B/C, B/D, E/F, and E/G) and one non-matching control

ZFN (E/C) (Fig. 1). We transfected the ZFN modules intoHEK293 cells and analyzed their cellular distribution. Immu-nofluorescence analysis revealed localization of ZFN modulesas expected mainly within the nucleus (Fig. 2A).

Next we established a cellular reporter system to monitorZFN-mediated genome editing of the mutated Ush1c in thegenomic context. For this, cDNA coding for the N-terminal partof the murine Ush1c gene product harmonin, including thep.R31X mutation, the first two PDZ domains (named after PSD-95, DLG, and ZO-1), and a C-terminal Myc-tag, was stablyintegrated at a single location in the genome of Flp in-HEK293cells. In parallel, Ush1c cells carrying the nonmutatedconstruct were established.

Customized ZFNs Have No Cytotoxicity

Since application of ZFNs is often associated with significantcytotoxicity,22 we examined potential side effects of ZFNs inthe p.R31X reporter system by measuring the cell viability inan XTT-based assay. For this, p.R31X cells were transfectedwith combinations of ZFN modules suitable to assemble activeZFN heterodimers, and 48 hours post transfection, the relativenumber of viable cells was determined (Fig. 2B). Transfectionof p.R31X cells with ZFN did not lead to a significant alterationin the number of viable cells compared with mock or emptyvector-transfected cells, indicating no cytotoxic side effects ofthe applied ZFNs.

DNA Cleavage Capability of Customized ZFN

To evaluate the cleavage capability of generated ZFNs, thep.R31X cell line was transfected with either matching ZFNs orthe control ZFN. Cells were harvested, genomic DNA waspurified, and semiquantitative PCR with a primer pair flankingthe putative ZFN cleavage site was performed (Fig. 2C).Efficient ZFN-mediated cleavage resulted in either a faint or acomplete lack of the PCR product as observed aftertransfection of ZFN module combinations A/C and E/F (Fig.2C). In contrast, the transfection of combinations A/D, B/C, E/G, and the control ZFN resulted in a PCR amplicon, indicatingthat these ZFNs are not able to cleave the target sequence.Transfections of ZFN B/D, however, led to varying results (Fig.2C). Based on these findings, we focused on the active ZFNs A/C and E/F in the following experiments as these seemed to bethe most efficient ZFNs.

Gene Repair Mediated by ZFN-Induced HR

Next we evaluated whether we can stimulate gene repair of thep.R31X mutation mediated by ZFN-induced HR. For thispurpose we co-transfected p.R31X cells with ZFN and a rescueplasmid that contains the nonmutated Ush1c sequencewithout any promoter. Cells were harvested, genomic DNAwas purified, and a PCR with a primer pair flanking themutation was performed. To omit amplification of the rescueplasmid, we designed a forward primer, binding within theCMV promoter located at the 50 end of the integrated p.R31Xconstruct in the Flp in-HEK293 cells, and a reverse primerspecific for the Ush1c sequence (Fig. 3A). Gene repair wasanalyzed by independent, complementary experiments. ThePCR products were digested with TseI, which only cleaves theUsh1c (GCT/GC) but not the mutated p.R31X sequence (GCT/GT). The digestion resulted in a single band (500 bp) in the un-transfected p.R31X cells and a double band (200 and 300 bp)in the Ush1c cells (Fig. 3B). TseI digestion after transfection ofmatching ZFN modules A/C and E/F resulted in clear doublebands (200 and 300 bp), showing an efficient gene repair ofthe p.R31X mutation mediated by ZFN (Fig. 3B). Transfection

4142 Overlack et al. IOVS, June 2012, Vol. 53, No. 7

of the control ZFN resulted in three bands (200, 300, and 500bp), indicating only weak repair efficiency of the mutatedsequence most likely owing to HR of the rescue DNA with themutated sequence.

To independently verify these results for repair by HR, wesequenced the undigested 500-bp PCR product, which wasamplified using a primer set composed of the CMV promoterforward primer and the Ush1c reverse primer. This ensuredamplification of the integrated genomic sequence and exclud-ed amplification of the rescue plasmid. Four (A/C) and five (E/F) PCR products after gene repair were sequenced. The resultsconfirmed that the generated PCR product after transfection ofmatching ZFN was a mix of Ush1c and p.R31X sequencesrepresented by the two peaks in the electropherogram (Fig.

FIGURE 3. Gene repair mediated by ZFN induced HR on the genomiclevel. (A) Stable transfected p.R31X cells were triple transfected withtwo ZFN modules and rescue plasmid. PCR was performed on genomicDNA with a primer pair (arrows) flanking the mutation (X). (B) TseIdigestion of PCR products. TseI cleaved only the Ush1c, not themutated p.R31X sequence. Fragment sizes: wild type, 200/300 bp;mutated p.R31X sequence, 500 bp. Transfection with matching ZFN A/C and E/F resulted in gene repair of p.R31X in the Ush1c sequence. (C)Sequencing of the undigested 500-bp PCR product showed a mix ofUsh1c and p.R31X sequences, with a higher peak for repaired Ush1c.This confirms an alteration of the mutated codon TGA to the wild typeCGG, generated from the rescue plasmid by HR, shown for ZFN E/F.Asterisks indicate exchanged nucleotides due to ZFN-mediated HR.

FIGURE 2. ZFN expression and evaluation in cell culture. (A) ZFNs aremainly localized in the nucleus of HEK293 cells, detected by anti-Flagantibody, exemplarily shown for ZFN module D. DAPI, nuclear DNAstaining; scale bar: 25 lm. (B) ZFN related cytotoxicity was analyzed bymeasurement of cell viability in XTT assay. Cells transfected withmatching or control ZFNs do not show an elevated rate of cell deathcompared with mock transfected cells or p.R31X cells transfected withthe empty vector. Addition of G418 was used as positive control for theassay. (C) In vitro cleavage of ZFN. Semiquantitative duplex PCR withgenomic DNA of p.R31X cells transfected with ZFN or un-transfectedwas performed with a primer pair (arrows) flanking the putative ZFNcleavage site (X). The two gels represent independent experiments.The faint or complete lack of the PCR product after transfection of A/Cand E/F indicated highly efficient DNA cleavage by these ZFNs. Incontrast, the transfection of combinations A/D, B/C, E/G, and thecontrol ZFN resulted in a more prominent PCR amplicon, indicatingthat these ZFNs are not able to cleave the target sequence efficiently.

Transfections of B/D led to varying results. The EcoRI digestion servedas a positive control for the cleavage, and GAPDH amplification servedas an internal PCR control.


3B; also, see Supplementary Material and Supplementary Fig.S1, http://www.iovs.org/lookup/suppl/doi:10.1167/iovs.12-9812/-/DCSupplemental). In all sequencing results, weobserved a higher peak for the sequence of the Ush1c siterepaired by HR, as shown for ZFN E/F and ZFN A/C (Fig. 3C;Supplementary Fig. S1, http://www.iovs.org/lookup/suppl/doi:10.1167/iovs.12-9812/-/DCSupplemental), respectively.These higher electropherogram peaks for the repairedsequence indicated higher abundance of the repaired nucleo-tides compared with mutated nucleotides, suggesting efficientgene repair of the p.R31X mutation. Furthermore, in oursequencings we did not observe any further base pairalterations besides the desired sequence repair (Supplementa-ry Fig. S1, http://www.iovs.org/lookup/suppl/doi:10.1167/iovs.12-9812/-/DCSupplemental). Although, we cannot com-pletely exclude non-homologous end joining (NHEJ), our dataindicate that HR is the preferred repair mechanism for DSBrepair.

ZFN-Induced Gene Correction by HR RecoversHarmonin Protein Expression

To demonstrate recovered expression of the wild type Ush1c

gene product harmonin on protein level we co-transfectedp.R31X cells with matching ZFNs and the rescue plasmid.Gene repair was analyzed by indirect immunofluorescenceusing anti-Myc antibody against the C-terminal tag of therecovered protein and Western blot analyses with antibodiesagainst harmonin (Fig. 4). No Myc-positive cells were detectedin un-transfected cells (Fig. 4A, upper panel). In contrast, co-transfection of matching ZFNs (A/C and E/F) with rescueplasmid resulted in Myc-positive cells (Fig. 4A). This clearlyindicated gene repair of the mutated Ush1c sequence resultingin recovered expression of the harmonin construct.

In Western blots of un-transfected p.R31X cells or p.R31Xcells transfected with the control ZFN, no anti-harmoninpositive band was detected, indicating that harmonin was notexpressed (Fig. 4B). As expected, anti-harmonin Western blotanalyses from lysates of Ush1c cells revealed a band of about 43kDa, corresponding to the calculated molecular weight of theharmonin construct (Fig. 4B). A band of the same molecularweight was detected in lysates of p.R31X cells co-transfectedwith matching ZFNs and rescue plasmid, demonstrating ZFN-induced restoration of harmonin (Fig. 4B).

DISCUSSION

Continual efforts have brought ZFN technology into the focusof faster and more efficient genome editing at selectedpredetermined sites in diverse eukaryotic species.7,11,12,23–25

So far, most of the reported successful applications of the ZFNtechnology have been for the introduction of targetedmodifications in the genome to create transgenic species.26–31

More recently, a series of key publications has emphasized thesignificant therapeutic potential of ZFNs, placing this technol-ogy into the spotlight in the field of gene-based therapy.32–34

In our present study, we introduced the ZFN technology usingcustomized ZFNs to induce HR for gene repair of a specificdisease-causing mutation responsible for a severe senso-neuronal degeneration, namely, the human Usher syndromeaffecting the inner ear and retina. Our data indicate that thedesigned ZFNs are suitable for gene targeting to induce HR tocorrect the p.R31X mutation in the Ush1c gene.

To ensure the ZFN-mediated introduction of DSBs at the siteof the mutation to be repaired, we have designed ZFNscontaining one ZF binding directly to the p.R31X-mutated site.Based on this strategy, we selected ZFs that did not always

target to GNN triplets, which were thought to be required forefficient binding of the ZF domain to the target DNA.35,36

Nevertheless, the efficient cleavage and the high biocompat-ibility of the generated ZFNs indicate that the ZF domaincomposition out of ZF exclusively recognizing GNN triplets isnot crucial, which is in line with previously reported data.27

The observed differences in the binding and cleavageefficiency of the tested ZFNs are probably owing to varyingsynergistic effects between adjacent ZFs as previously dis-cussed.37

In the present study, we provide proof-of-concept resultsthat designed ZFNs mediate gene correction of a mutated USHgene efficiently through the introduction of site-specific DSBsat the mutation. Cells have developed two, in principle,different mechanisms to repair DSBs: NHEJ and homology-directed repair.38 Since in NHEJ DNA ends are re-ligatedwithout any use of homology, NHEJ is an error-pronemechanism, often introducing mutagenic changes.39 In con-

FIGURE 4. Gene repair mediated by ZFN induced HR on protein level.(A) Indirect immunofluorescence analysis of p.R31X cells revealedMyc-positive cells after transfection of ZFN A/C and E/F together withthe rescue plasmid. No Myc labeling was detected in un-transfectedcells. DAPI, nuclear DNA staining; scale bar: 25 lm. (B) Western blotanalysis of p.R31X cells, either un-transfected or transfected with ZFNand rescue plasmid. Cell lysates were subjected to Western blot withanti-harmonin and anti-tubulin antibodies (loading control). Harmonin-PDZ2-Myc expression (~43 kDa) was detected by anti-H3 in Ush1c

wild type cells and in pR31X cells transfected with ZFN A/C and E/F.


trast, HR is a ‘‘copy and paste’’ mechanism, using a DNAtemplate for the correct repair of DSBs preferred fortherapeutic gene correction.40 In our study, we activate HRproviding the rescue DNA as the correct template for repair ofZFN-induced DSB. Our results provide several lines of evidencefor the site-specific gene correction of the USH disease-causingp.R31X mutation by HR. The present cleavage assay revealedintroduction of the wild type sequence into the mutant p.R31Xcell line after application of matching ZFNs in concert with therescue DNA. Sequence analysis revealed the rescue-specificgenotype, indicating that rescue DNA was used as a template tofill up the gap introduced by ZFN-induced DSB. Finally, wedemonstrated the recovery of harmonin protein expression byimmunocytochemistry and Western blot analysis.

One significant drawback for the therapeutic application ofthe ZFN technology is the frequently reported toxic sideeffects of ZFNs introduced by off-target DNA cleavage.41–43

Nevertheless, based on the results of cell viability tests, we didnot observe cytotoxicity of the customized p.R31X ZFNs,indicating no obvious off-target effects. Our findings are in linewith previous publications.14,30

We customized our ZFNs to target the Ush1c gene, with thefuture goal of correcting mutations causing human USHaffecting the sensory neurons in the inner ear and the retina.Although the mechanism of HR has been well-established forcycling cells,44 only recently was HR identified as a DSB repairpathway in terminally differentiated neurons as photoreceptorcells.45 The authors of the study state that about 15% of DSB inrod photoreceptor cells are repaired by applying homology-based repair mechanisms, even without adding homologueDNA sequences.45 This further qualifies our ZFN-based repairapproach for correcting USH genes in the eye and ear. Thepossibility of repairing a disease-causing mutation in onetherapeutic intervention instead of continuously treating thesymptoms holds great advantages especially for eye diseases.Because of an effective blood-retina-barrier, systemic applica-tion of therapeutics to the retina is very challenging.5

Gene addition by viral vectors is traded as a promisingstrategy for the treatment of USH and is in preclinical trials forUSH1B46 (in the public domain; www.oxfordbiomedica.co.uk).Since USH genes frequently are very large and/or have varioussplice variants, gene addition approaches are very challenging.More recently, we succeeded in introducing translationalreadthrough-inducing drugs as a therapeutic intervention forUSH1 nonsense mutations.47,48 However, this drug treatment isonly suitable for ~12% of USH mutations,49 and lifelongmedication is needed. The presented strategy of ZFNtechnology has apparent advantages over established gene-based therapy approaches5 as (1) the repaired gene stays undercontrol of its endogenous promoter, so expression level,splicing, and spatio-temporal expression remain unaffected; (2)no lifelong medication is necessary; (3) the size of the geneaffected is not an issue; and (4) the type of the disease-causingmutation is not an issue.

Taken together, the good biocompatibility of our ZFNs andrepair of the p.R31X mutation in our cell culture systemprovide evidence that ZFN-induced HR is suitable for thetreatment of inherited genetic disorders, such as USH or relatedretinal ciliopathies, to cure the affected terminally differenti-ated sensory neurons in the retina. Conclusively, the ZFNtechnology holds an extensive potential for the treatment ofknown mutations in the direction of personalized therapystrategies.

Acknowledgments

The authors thank Philipp Trojan and Michiel van Wyk for criticalreading of this manuscript and Marcel Alavi for help with

experimental design. The authors also thank Keith Joung forproviding the zinc-finger cDNA and plasmids.

References

1. Reiners J, Nagel-Wolfrum K, Jurgens K, Marker T, Wolfrum U.Molecular basis of human Usher syndrome: deciphering themeshes of the Usher protein network provides insights intothe pathomechanisms of the Usher disease. Exp Eye Res. 2006;83:97–119.

2. Wolfrum U. Protein networks related to the Usher syndromegain insights in the molecular basis of the disease. In: Satpal A,ed. Usher Syndrome: Pathogenesis, Diagnosis and Therapy.New York, NY: Nova Science Publishers, Inc.; 2011:51–73.

3. Verpy E, Leibovici M, Zwaenepoel I, et al. A defect inharmonin, a PDZ domain-containing protein expressed in theinner ear sensory hair cells, underlies Usher syndrome type1C. Nat Genet. 2000;26:51–55.

4. den Hollander AI, Black A, Bennett J, Cremers FP. Lighting acandle in the dark: advances in genetics and gene therapy ofrecessive retinal dystrophies. J Clin Invest. 2010;120:3042–3053.

5. Overlack N, Goldmann T, Wolfrum U, Nagel-Wolfrum K.Current therapeutic strategies for human Usher syndrome. In:Ahuja S, ed. Usher Syndrome: Pathogenesis, Diagnosis and

Therapy. New York, NY: Nova Science Publishers, Inc.; 2011:377–395.

6. Nagel-Wolfrum K, Becker M, Goldmann T, Muller C, Vetter J,Wolfrum U. USH1C transcripts and harmonin protein expres-sion in human retina [abstract]. Invest Ophthalmol Vis Sci.2011;52:45.

7. Urnov FD, Rebar EJ, Holmes MC, Zhang HS, Gregory PD.Genome editing with engineered zinc finger nucleases. Nat

Rev Genet. 2010;11:636–646.

8. Orr-Weaver TL, Szostak JW, Rothstein RJ. Yeast transformation:a model system for the study of recombination. Proc Natl

Acad Sci U S A. 1981;78:6354–6358.

9. Porteus MH, Baltimore D. Chimeric nucleases stimulate genetargeting in human cells. Science. 2003;300:763.

10. Rouet P, Smih F, Jasin M. Introduction of double-strand breaksinto the genome of mouse cells by expression of a rare-cuttingendonuclease. Mol Cell Biol. 1994;14:8096–8106.

11. Bibikova M, Carroll D, Segal DJ, et al. Stimulation ofhomologous recombination through targeted cleavage bychimeric nucleases. Mol Cell Biol. 2001;21:289–297.

12. Carroll D. Progress and prospects: zinc-finger nucleases asgene therapy agents. Gene Ther. 2008;15:1463–1468.

13. Lombardo A, Genovese P, Beausejour CM, et al. Gene editingin human stem cells using zinc finger nucleases and integrase-defective lentiviral vector delivery. Nat Biotechnol. 2007;25:1298–1306.

14. Greenwald DL, Cashman SM, Kumar-Singh R. Engineered zincfinger nuclease-mediated homologous recombination of thehuman rhodopsin gene. Invest Ophthalmol Vis Sci. 2010;51:6374–6380.

15. Rahman SH, Maeder ML, Joung JK, Cathomen T. Zinc-fingernucleases for somatic gene therapy: the next frontier. Hum

Gene Ther. 2011;22:925–933.

16. Sander JD, Zaback P, Joung JK, Voytas DF, Dobbs D. Zinc fingertargeter (ZiFiT): an engineered zinc finger/target site designtool. Nucleic Acids Res. 2007;35:W599–W605.

17. Wright DA, Thibodeau-Beganny S, Sander JD, et al. Standard-ized reagents and protocols for engineering zinc fingernucleases by modular assembly. Nat Protoc. 2006;1:1637–1652.

18. Reiners J, Reidel B, El-Amraoui A, et al. Differential distributionof harmonin isoforms and their possible role in Usher-1


protein complexes in mammalian photoreceptor cells. Invest

Ophthalmol Vis Sci. 2003;44:5006–5015.

19. Overlack N, Maerker T, Latz M, Nagel-Wolfrum K, Wolfrum U.SANS (USH1G) expression in developing and mature mamma-lian retina. Vision Res. 2008;48:400–412.

20. Nagel-Wolfrum K, Buerger C, Wittig I, Butz K, Hoppe-Seyler F,Groner B. The interaction of specific peptide aptamers withthe DNA binding domain and the dimerization domain of thetranscription factor Stat3 inhibits transactivation and inducesapoptosis in tumor cells. Mol Cancer Res. 2004;2:170–182.

21. Maerker T, van Wijk E, Overlack N, et al. A novel Usher proteinnetwork at the periciliary reloading point between moleculartransport machineries in vertebrate photoreceptor cells. Hum

Mol Genet. 2008;17:71–86.

22. Szczepek M, Brondani V, Buchel J, Serrano L, Segal DJ,Cathomen T. Structure-based redesign of the dimerizationinterface reduces the toxicity of zinc-finger nucleases. Nat

Biotechnol. 2007;25:786–793.

23. Morton J, Davis MW, Jorgensen EM, Carroll D. Induction andrepair of zinc-finger nuclease-targeted double-strand breaks inCaenorhabditis elegans somatic cells. Proc Natl Acad Sci U S

A. 2006;103:16370–16375.

24. Cathomen T, Joung JK. Zinc-finger nucleases: the nextgeneration emerges. Mol Ther. 2008;16:1200–1207.

25. Doyon Y, McCammon JM, Miller JC, et al. Heritable targetedgene disruption in zebrafish using designed zinc-fingernucleases. Nat Biotechnol. 2008;26:702–708.

26. Geurts AM, Cost GJ, Freyvert Y, et al. Knockout rats viaembryo microinjection of zinc-finger nucleases. Science. 2009;325:433.

27. Takasu Y, Kobayashi I, Beumer K, et al. Targeted mutagenesisin the silkworm Bombyx mori using zinc finger nucleasemRNA injection. Insect Biochem Mol Biol. 2010;40:759–765.

28. Baylis HA, Vazquez-Manrique RP. Reverse genetic strategies inCaenorhabditis elegans: towards controlled manipulation ofthe genome. ScientificWorldJournal. 2011;11:1394–1410.

29. Even-Faitelson L, Samach A, Melamed-Bessudo C, Avivi-Ragolsky N, Levy AA. Localized egg-cell expression of effectorproteins for targeted modification of the Arabidopsis genome.Plant J. 2011;68:929–937.

30. Hauschild J, Petersen B, Santiago Y, et al. Efficient generationof a biallelic knockout in pigs using zinc-finger nucleases. Proc

Natl Acad Sci U S A. 2011;108:12013–12017.

31. McCammon JM, Doyon Y, Amacher SL. Inducing high rates oftargeted mutagenesis in zebrafish using zinc finger nucleases(ZFNs). Methods Mol Biol. 2011;770:505–527.

32. Li H, Haurigot V, Doyon Y, et al. In vivo genome editingrestores haemostasis in a mouse model of haemophilia.Nature. 2011;475:217–221.

33. Soldner F, Laganiere J, Cheng AW, et al. Generation of isogenicpluripotent stem cells differing exclusively at two early onsetParkinson point mutations. Cell. 2011;146:318–331.

34. Zou J, Luo L, Shen Z, et al. Whirlin replacement restores theformation of the USH2 protein complex in whirlin knockout

photoreceptors. Invest Ophthalmol Vis Sci. 2011;52:2343–2351.

35. Beumer K, Bhattacharyya G, Bibikova M, Trautman JK, CarrollD. Efficient gene targeting in Drosophila with zinc-fingernucleases. Genetics. 2006;172:2391–2403.

36. Ramirez CL, Foley JE, Wright DA, et al. Unexpected failurerates for modular assembly of engineered zinc fingers. Nat

Methods. 2008;5:374–375.

37. Isalan M, Choo Y, Klug A. Synergy between adjacent zincfingers in sequence-specific DNA recognition. Proc Natl Acad

Sci U S A. 1997;94:5617–5621.

38. Pardo B, Gomez-Gonzalez B, Aguilera A. DNA repair inmammalian cells: DNA double-strand break repair: how tofix a broken relationship. Cell Mol Life Sci. 2009;66:1039–1056.

39. Jensen NM, Dalsgaard T, Jakobsen M, et al. An update ontargeted gene repair in mammalian cells: methods andmechanisms. J Biomed Sci. 2011;18:10.

40. Kramer O, Klausing S, Noll T. Methods in mammalian cell lineengineering: from random mutagenesis to sequence-specificapproaches. Appl Microbiol Biotechnol. 2010;88:425–436.

41. Pruett-Miller SM, Reading DW, Porter SN, Porteus MH.Attenuation of zinc finger nuclease toxicity by small-moleculeregulation of protein levels. PLoS Genet. 2009;5:e1000376.Avai lable at: http://www.plosgenetics.org/art icle/info%3Adoi%2F10.1371%2Fjournal.pgen.1000376. AccessedMarch 3, 2009.

42. Cornu TI, Cathomen T. Quantification of zinc finger nuclease-associated toxicity. Methods Mol Biol. 2010;649:237–245.

43. Gupta A, Meng X, Zhu LJ, Lawson ND, Wolfe SA. Zinc fingerprotein-dependent and -independent contributions to the invivo off-target activity of zinc finger nucleases. Nucleic Acids

Res. 2011;39:381–392.

44. Kass EM, Jasin M. Collaboration and competition betweenDNA double-strand break repair pathways. FEBS Lett. 2010;584:3703–3708.

45. Chan F, Hauswirth WW, Wensel TG, Wilson JH. Efficientmutagenesis of the rhodopsin gene in rod photoreceptorneurons in mice. Nucleic Acids Res. 2011;39:5955–5966.

46. Hashimoto T, Gibbs D, Lillo C, et al. Lentiviral genereplacement therapy of retinas in a mouse model for Ushersyndrome type 1B. Gene Ther. 2007;14:584–594.

47. Goldmann T, Rebibo-Sabbah A, Overlack N, et al. Beneficialread-through of a USH1C nonsense mutation by designedaminoglycoside NB30 in the retina. Invest Ophthalmol Vis Sci.2010;51:6671–6680.

48. Goldmann T, Overlack N, Wolfrum U, Nagel-Wolfrum K.PTC124-mediated translational readthrough of a nonsensemutation causing Usher syndrome type 1C. Hum Gene Ther.2011;22:537–547.

49. Kellermayer R. Translational readthrough induction of patho-genic nonsense mutations. Eur J Med Genet. 2006;49:445–450.


A

FokI

FokIG G G G G G G G G G G G G GA A A A A A A A A A AC C C C C C C C CT T T T T T T T T T T T T T T

ZFN E/F

FokI

FokIG G G G G G G G G G GA A A A A A A A A A AC C C C C C C C CT T T T T T T T T T T T T T

BZFN A/C

** **

**

ATG

C/1 C/3C/2

A/3 A/2 A/1

F/1 F/2 F/3

E/3 E/2 E/1

T A T A

NN

NN

Supplementary Figure S1. ZFN induced gene repair does not alter bp sequences surrounding the p.R31X mutation. (A)

(B)

Sequence alignment of wild type Ush1c sequence with sequencing results of gene repair after transfection of ZFN A/C and E/F. Weobserved no further bp alterations 5´and 3´of the desired sequence repair (bp 91 and bp 93). Dotted line: stably integrated vectorsequences; solid line: Ush1c sequence up to bp 240; orange bar: ZFN A/C binding site (bp 74-96); brown bar: ZFN E/F binding site (bp87-109); ATG: Ush1c start codon; asterisks/N: desired sequence repair of the p.R31X mutation. Sequencing electropherogram ofrepaired Ush1c sequence after transfection of ZFN A/C (orange) or E/F (brown). Binding sites of ZF-domains (grey ovals) and spacerfor FokI dimerization are indicated.

gene repair of an usher syndrome causing mutation by zinc-finger nuclease … · 2019-01-10 ·...

Documents